def __init__(self, checkpoint_path, device,
img_mean=np.array([128, 128, 128], dtype=np.float32),
img_scale=np.float32(1/255),
use_tensorrt=False):
from models.with_mobilenet import PoseEstimationWithMobileNet
from modules.load_state import load_state
self.img_mean = img_mean
self.img_scale = img_scale
self.device = 'cpu'
if device != 'CPU':
if torch.cuda.is_available():
self.device = torch.device('cuda:0')
else:
print('No CUDA device found, inferring on CPU')
net = PoseEstimationWithMobileNet()
checkpoint = torch.load(checkpoint_path, map_location='cpu')
if use_tensorrt:
from torch2trt import TRTModule
net = TRTModule()
net.load_state_dict(checkpoint)
else:
load_state(net, checkpoint)
net = net.to(self.device)
net.eval()
self.net = net
def init_pose(checkpoint_path):
net = PoseEstimationWithMobileNet()
checkpoint = torch.load(checkpoint_path, map_location='cpu')
load_state(net, checkpoint)
net.eval()
net = net.cuda()
env = [net, []]
return None, env
def openpose_to_jit():
x = torch.randn(1,3,256,456)
net = PoseEstimationWithMobileNet().cpu()
checkpoint = torch.load(r'.\weights\checkpoint_iter_370000.pth', map_location='cpu')
load_state(net, checkpoint)
net.eval()
net(x)
script_model = torch.jit.trace(net, x)
script_model.save('test.jit')
def __init__(self):
self.name = 'OpenPose'
net = PoseEstimationWithMobileNet()
checkpoint = torch.load('./checkpoint_iter_370000.pth',
map_location='cpu')
load_state(net, checkpoint)
self.net = net.eval()
if envars.USE_GPU():
self.net = self.net.cuda()
self.stride = 8
self.upsample_ratio = 4
self.height_size = 256
self.kpt_names = [
'nose', 'neck', 'r_sho', 'r_elb', 'r_wri', 'l_sho', 'l_elb',
'l_wri', 'r_hip', 'r_knee', 'r_ank', 'l_hip', 'l_knee', 'l_ank',
'r_eye', 'l_eye', 'r_ear', 'l_ear'
]
self.connections = [('nose', 'r_eye'), ('r_eye', 'r_ear'),
('nose', 'l_eye'), ('l_eye', 'l_ear'),
('nose', 'neck'), ('neck', 'r_sho'),
('r_sho', 'r_elb'), ('r_elb', 'r_wri'),
('neck', 'l_sho'), ('l_sho', 'l_elb'),
('l_elb', 'l_wri'), ('neck', 'r_hip'),
('r_hip', 'r_knee'), ('r_knee', 'r_ank'),
('neck', 'l_hip'), ('l_hip', 'l_knee'),
('l_knee', 'l_ank')]
def init(cpu = False):
net = PoseEstimationWithMobileNet()
checkpoint_path = "checkpoint_iter_370000.pth"
checkpoint = torch.load(checkpoint_path, map_location='cpu') #load the existing model
load_state(net, checkpoint)
net = net.eval()
if not cpu:
net = net.cuda()
return net
def run_demo(args, image_provider, height_size, cpu, track, smooth):
net = PoseEstimationWithMobileNet()
checkpoint = torch.load(args.checkpoint_path, map_location='cpu')
load_state(net, checkpoint)
net = net.eval()
if not cpu:
net = net.cuda()
stride = 8
upsample_ratio = 4
num_keypoints = Pose.num_kpts
previous_poses = []
delay = 33
for d in image_provider:
img, image_name = d["image"], d["image_name"]
orig_img = img.copy()
heatmaps, pafs, scale, pad = infer_fast(net, img, height_size, stride,
upsample_ratio, cpu)
total_keypoints_num = 0
all_keypoints_by_type = []
for kpt_idx in range(num_keypoints): # 19th for bg
total_keypoints_num += extract_keypoints(heatmaps[:, :, kpt_idx],
all_keypoints_by_type,
total_keypoints_num)
pose_entries, all_keypoints = group_keypoints(all_keypoints_by_type,
pafs,
demo=True)
for kpt_id in range(all_keypoints.shape[0]):
all_keypoints[kpt_id, 0] = (all_keypoints[kpt_id, 0] * stride /
upsample_ratio - pad[1]) / scale
all_keypoints[kpt_id, 1] = (all_keypoints[kpt_id, 1] * stride /
upsample_ratio - pad[0]) / scale
keypoints_out = []
for n in range(len(pose_entries)):
if len(pose_entries[n]) == 0:
continue
pose_keypoints = np.ones((num_keypoints, 3), dtype=np.float32) * -1
for kpt_id in range(num_keypoints):
if pose_entries[n][kpt_id] != -1.0: # keypoint was found
pose_keypoints[kpt_id, 0] = int(
all_keypoints[int(pose_entries[n][kpt_id]), 0])
pose_keypoints[kpt_id, 1] = int(
all_keypoints[int(pose_entries[n][kpt_id]), 1])
pose_keypoints[kpt_id,
2] = 0.94 if pose_keypoints[kpt_id,
0] != -1 else 0
keypoints_out.append(pose_keypoints)
save_json(image_name, keypoints_out, args)
class LightBody(object):
def __init__(self, model_path):
self.model = PoseEstimationWithMobileNet()
checkpoint = torch.load(model_path, map_location='cpu')
load_state(self.model, checkpoint)
self.model = self.model.eval()
self.model = self.model.cuda()
def __call__(self, img, height_size=256):
stride = 8
upsample_ratio = 4
num_keypoints = Pose.num_kpts
orig_img = img.copy()
heatmaps, pafs, scale, pad = infer_fast(self.model, img, height_size,
stride, upsample_ratio, False)
total_keypoints_num = 0
all_keypoints_by_type = []
for kpt_idx in range(num_keypoints): # 19th for bg
total_keypoints_num += extract_keypoints(heatmaps[:, :, kpt_idx],
all_keypoints_by_type,
total_keypoints_num)
pose_entries, all_keypoints = group_keypoints(all_keypoints_by_type,
pafs)
for kpt_id in range(all_keypoints.shape[0]):
all_keypoints[kpt_id, 0] = (all_keypoints[kpt_id, 0] * stride /
upsample_ratio - pad[1]) / scale
all_keypoints[kpt_id, 1] = (all_keypoints[kpt_id, 1] * stride /
upsample_ratio - pad[0]) / scale
current_poses = []
for n in range(len(pose_entries)):
if len(pose_entries[n]) == 0:
continue
pose_keypoints = np.ones((num_keypoints, 2), dtype=np.int32) * -1
for kpt_id in range(num_keypoints):
if pose_entries[n][kpt_id] != -1.0: # keypoint was found
pose_keypoints[kpt_id, 0] = int(
all_keypoints[int(pose_entries[n][kpt_id]), 0])
pose_keypoints[kpt_id, 1] = int(
all_keypoints[int(pose_entries[n][kpt_id]), 1])
pose = Pose(pose_keypoints, pose_entries[n][18])
current_poses.append(pose)
for pose in current_poses:
pose.draw(img)
img = cv2.addWeighted(orig_img, 0.6, img, 0.4, 0)
cv2.imwrite("/data1/qilei_chen/DEVELOPMENTS/test1.jpg", img)
return current_poses
def _pose_dect_init(self, device):
"""Initialize the pose detection model.
Arguments:
device {torch.device}: device to implement the models on.
Returns:
PoseEstimationWithMobileNet: initialized OpenPose model.
"""
weight_path = self.__params.pose_weights
model = PoseEstimationWithMobileNet()
weight = torch.load(weight_path, map_location='cpu')
load_state(model, weight)
model = model.eval()
if device.type != 'cpu':
model = model.cuda()
return model
def callback(self, data):
try:
cv_image = self.bridge.imgmsg_to_cv2(data, "bgr8")
except CvBridgeError as e:
print(e)
## Rescale Image size
rescale_factor = 1
width = int(cv_image.shape[1] * rescale_factor)
height = int(cv_image.shape[0] * rescale_factor)
dim = (width, height)
resized_img = cv2.resize(cv_image, dim)
net = PoseEstimationWithMobileNet()
checkpoint = torch.load(
"/home/zheng/lightweight-human-pose-estimation.pytorch/checkpoint_iter_370000.pth",
map_location='cpu')
load_state(net, checkpoint)
height_size = 256
net = net.eval()
net = net.cuda()
net.eval()
stride = 8
upsample_ratio = 4
num_keypoints = Pose.num_kpts
previous_poses = []
delay = 33
# img = cv2.imread("/home/zheng/lightweight-human-pose-estimation.pytorch/data/image_1400.jpg")
img = asarray(cv_image)
orig_img = img
heatmaps, pafs, scale, pad = infer_fast(net,
img,
height_size,
stride,
upsample_ratio,
cpu="store_true")
total_keypoints_num = 0
all_keypoints_by_type = []
for kpt_idx in range(num_keypoints): # 19th for bg
total_keypoints_num += extract_keypoints(heatmaps[:, :, kpt_idx],
all_keypoints_by_type,
total_keypoints_num)
pose_entries, all_keypoints = group_keypoints(all_keypoints_by_type,
pafs,
demo=True)
for kpt_id in range(all_keypoints.shape[0]):
all_keypoints[kpt_id, 0] = (all_keypoints[kpt_id, 0] * stride /
upsample_ratio - pad[1]) / scale
all_keypoints[kpt_id, 1] = (all_keypoints[kpt_id, 1] * stride /
upsample_ratio - pad[0]) / scale
current_poses = []
## Collect all keypoint in numpy array to send it to Ros"
pose_keypoints_ros_data = np.zeros(16)
my_array_for_publishing = Float32MultiArray()
####
pose_keypoints = np.ones((num_keypoints, 2), dtype=np.int32) * -1
for kpt_id in range(8):
if pose_entries[0][kpt_id] != -1.0: # keypoint was found
pose_keypoints[kpt_id, 0] = int(
all_keypoints[int(pose_entries[0][kpt_id]), 0])
pose_keypoints[kpt_id, 1] = int(
all_keypoints[int(pose_entries[0][kpt_id]), 1])
pose = Pose(pose_keypoints, pose_entries[0][18])
current_poses.append(pose)
pose_keypoints_ros_data[2 * kpt_id] = pose.keypoints[kpt_id][0]
pose_keypoints_ros_data[2 * kpt_id + 1] = pose.keypoints[kpt_id][1]
for pose in current_poses:
pose.draw(img)
img = cv2.addWeighted(orig_img, 0.6, img, 0.4, 0)
my_array_for_publishing.data = [
pose_keypoints_ros_data[0],
pose_keypoints_ros_data[1],
pose_keypoints_ros_data[2],
pose_keypoints_ros_data[3],
pose_keypoints_ros_data[4],
pose_keypoints_ros_data[5],
pose_keypoints_ros_data[6],
pose_keypoints_ros_data[7],
pose_keypoints_ros_data[8],
pose_keypoints_ros_data[9],
pose_keypoints_ros_data[10],
pose_keypoints_ros_data[11],
pose_keypoints_ros_data[12],
pose_keypoints_ros_data[13],
pose_keypoints_ros_data[14],
pose_keypoints_ros_data[15],
]
# cv2.imshow('Lightweight Human Pose Estimation Python Demo', img)
self.image_pub.publish(self.bridge.cv2_to_imgmsg(img, "bgr8"))
self.keypts_pub.publish(my_array_for_publishing)
# cv2.imwrite('/home/zheng/Bureau/image_1400_key.jpg',img)
cv2.waitKey(2)
setting.text3.SetLabelText("{} : {}".format(mainstudentcap.username,
mainstudentcap.score))
setting.text2.SetLabelText("{} : {}".format(secondstudentcap.username,
secondstudentcap.score))
setting.text1.SetLabelText("{} : {}".format(thirdstudentcap.username,
thirdstudentcap.score))
if __name__ == '__main__':
net = PoseEstimationWithMobileNet()
checkpoint = torch.load("checkpoint\\checkpoint_iter_370000.pth",
map_location='cpu')
load_state(net, checkpoint)
net = net.cuda()
net = net.eval()
capture = None
capture = cv2.VideoCapture("data\\HsinDance.mp4")
capture.set(cv2.CAP_PROP_FRAME_WIDTH, 1280)
capture.set(cv2.CAP_PROP_FRAME_HEIGHT, 720)
capture4 = cv2.VideoCapture("data\\IronmanDance24.mp4")
capture4.set(cv2.CAP_PROP_FRAME_WIDTH, 1280)
capture4.set(cv2.CAP_PROP_FRAME_HEIGHT, 720)
screen_width = user32.GetSystemMetrics(0)
screen_height = user32.GetSystemMetrics(1)
capture2 = cv2.VideoCapture(0)
capture2.set(cv2.CAP_PROP_FRAME_WIDTH, 1280)
capture2.set(cv2.CAP_PROP_FRAME_HEIGHT, 720)
humans.append(human)
#socket.send_string(json.dumps(humans))
img = cv2.addWeighted(orig_img, 0.8, img, 0.8, 0)
cv2.imshow('Human Pose', img)
key = cv2.waitKey(1)
if key == 27: # esc
return
if __name__ == '__main__':
net = PoseEstimationWithMobileNet()
checkpoint = torch.load("checkpoint.pth.tar", map_location='cpu')
load_state(net, checkpoint)
net.eval()
#example = torch.randn(1, 3, 256, 456)
#traced_script_module = torch.jit.script(net, example)
#traced_script_module.save('human-pose.pt')
#x = torch.rand(1, 3, 256, 456)
#sm = torch.jit.trace(net,[Variable(x)])
#sm.save("human-pose.pt")
frame_provider = VideoCapture(0)
frame_provider.start()
run_demo(net, frame_provider, 256, "cuda")
class PoseDetector(object):
def __init__(self, model_path: str):
self.device = torch.device(
"cuda" if torch.cuda.is_available() else "cpu")
self.net = PoseEstimationWithMobileNet()
self.net.to(self.device)
checkpoint = torch.load(model_path)
load_state(self.net, checkpoint)
self.net.eval()
self.image = None
self.avg_heatmap = None
self.avg_paf = None
self.track = True
self.stride = 8
self.upsample_ratio = 4
self.height_size = 256
self.smooth = 1
self.num_keypoints = Pose.num_kpts
def __inference(self, image, multiscale=False):
img = image.copy()
base_height = 368
scales = [1]
if multiscale:
scales = [0.5, 1.0, 1.5, 2.0]
stride = 8
normed_img = self.__normalize(img)
height, width, _ = normed_img.shape
scales_ratios = [
scale * base_height / float(height) for scale in scales
]
avg_heatmap = np.zeros((height, width, 19), dtype=np.float32)
avg_paf = np.zeros((height, width, 38), dtype=np.float32)
for ratio in scales_ratios:
scaled_img = cv2.resize(normed_img, (0, 0),
fx=ratio,
fy=ratio,
interpolation=cv2.INTER_CUBIC)
min_dims = [base_height, max(scaled_img.shape[1], base_height)]
padded_img, pad = self.__pad_width(scaled_img, stride, min_dims)
tensor_img = torch.from_numpy(padded_img).permute(
2, 0, 1).unsqueeze(0).float().to(self.device)
stages_output = self.net(tensor_img)
stage2_heatmap = stages_output[-2]
heatmap = np.transpose(stage2_heatmap.squeeze().cpu().data.numpy(),
(1, 2, 0))
heatmap = cv2.resize(heatmap, (0, 0),
fx=stride,
fy=stride,
interpolation=cv2.INTER_CUBIC)
heatmap = heatmap[pad[0]:heatmap.shape[0] - pad[2],
pad[1]:heatmap.shape[1] - pad[3]:, :]
heatmap = cv2.resize(heatmap, (width, height),
interpolation=cv2.INTER_CUBIC)
self.avg_heatmap = avg_heatmap + heatmap / len(scales_ratios)
stage2_paf = stages_output[-1]
paf = np.transpose(stage2_paf.squeeze().cpu().data.numpy(),
(1, 2, 0))
paf = cv2.resize(paf, (0, 0),
fx=stride,
fy=stride,
interpolation=cv2.INTER_CUBIC)
paf = paf[pad[0]:paf.shape[0] - pad[2],
pad[1]:paf.shape[1] - pad[3], :]
paf = cv2.resize(paf, (width, height),
interpolation=cv2.INTER_CUBIC)
self.avg_paf = avg_paf + paf / len(scales_ratios)
def __inference_fast(self, img, net_input_height_size, stride,
upsample_ratio):
height, width, _ = img.shape
self.scale = net_input_height_size / height
scaled_img = cv2.resize(img, (0, 0),
fx=self.scale,
fy=self.scale,
interpolation=cv2.INTER_LINEAR)
scaled_img = self.__normalize(scaled_img)
min_dims = [
net_input_height_size,
max(scaled_img.shape[1], net_input_height_size)
]
padded_img, self.pad = self.__pad_width(scaled_img, stride, min_dims)
tensor_img = torch.from_numpy(padded_img).permute(
2, 0, 1).unsqueeze(0).float()
tensor_img = tensor_img.to(self.device)
stages_output = self.net(tensor_img)
stage2_heatmaps = stages_output[-2]
heatmaps = np.transpose(stage2_heatmaps.squeeze().cpu().data.numpy(),
(1, 2, 0))
self.avg_heatmap = cv2.resize(heatmaps, (0, 0),
fx=upsample_ratio,
fy=upsample_ratio,
interpolation=cv2.INTER_CUBIC)
stage2_pafs = stages_output[-1]
pafs = np.transpose(stage2_pafs.squeeze().cpu().data.numpy(),
(1, 2, 0))
self.avg_paf = cv2.resize(pafs, (0, 0),
fx=upsample_ratio,
fy=upsample_ratio,
interpolation=cv2.INTER_CUBIC)
total_keypoints_num = 0
all_keypoints_by_type = []
for kpt_idx in range(self.num_keypoints): # 19th for bg
total_keypoints_num += extract_keypoints(
self.avg_heatmap[:, :, kpt_idx], all_keypoints_by_type,
total_keypoints_num)
self.pose_entries, self.all_keypoints = group_keypoints(
all_keypoints_by_type, self.avg_paf)
for kpt_id in range(self.all_keypoints.shape[0]):
self.all_keypoints[kpt_id, 0] = \
int((self.all_keypoints[kpt_id, 0] * self.stride / self.upsample_ratio - self.pad[1]) / self.scale)
self.all_keypoints[kpt_id, 1] = \
int((self.all_keypoints[kpt_id, 1] * self.stride / self.upsample_ratio - self.pad[0]) / self.scale)
def visualize_prediction(self, image):
orig_img = image.copy()
if not np.array_equal(self.image, image):
self.image = image
self.__inference_fast(self.image, self.height_size, self.stride,
self.upsample_ratio)
current_poses = []
for n in range(len(self.pose_entries)):
if len(self.pose_entries[n]) == 0:
continue
pose_keypoints = np.ones(
(self.num_keypoints, 2), dtype=np.int32) * -1
for kpt_id in range(self.num_keypoints):
if self.pose_entries[n][kpt_id] != -1.0: # keypoint was found
pose_keypoints[kpt_id, 0] = int(
self.all_keypoints[int(self.pose_entries[n][kpt_id]),
0])
pose_keypoints[kpt_id, 1] = int(
self.all_keypoints[int(self.pose_entries[n][kpt_id]),
1])
pose = Pose(pose_keypoints, self.pose_entries[n][18])
current_poses.append(pose)
# if self.track:
# previous_poses = []
# track_poses(previous_poses, current_poses, smooth=smooth)
# previous_poses = current_poses
for pose in current_poses:
pose.draw(image)
image = cv2.addWeighted(orig_img, 0.6, image, 0.4, 0)
# plt.imshow(image)
# plt.show()
return image
def __get_auto_grading_outputs(self, image):
if not np.array_equal(self.image, image):
self.image = image
self.__inference_fast(self.image, self.height_size, self.stride,
self.upsample_ratio)
return self.all_keypoints, self.pose_entries
def __visualize_prediction_slow(self, image):
if not np.array_equal(self.image, image):
self.image = image
self.__inference(self.image)
total_keypoints_num = 0
all_keypoints_by_type = []
for kpt_idx in range(18): # 19th for bg
total_keypoints_num += extract_keypoints(
self.avg_heatmap[:, :, kpt_idx], all_keypoints_by_type,
total_keypoints_num)
pose_entries, all_keypoints = group_keypoints(all_keypoints_by_type,
self.avg_paf)
coco_keypoints, scores = convert_to_coco_format(
pose_entries, all_keypoints)
for keypoints in coco_keypoints:
for idx in range(len(keypoints) // 3):
cv2.circle(
image,
(int(keypoints[idx * 3]), int(keypoints[idx * 3 + 1])), 3,
(255, 0, 255), -1)
plt.imshow(image)
plt.show()
def __get_auto_grading_outputs_slow(self, image):
if not np.array_equal(self.image, image):
self.image = image
self.__inference(self.image)
all_peaks = []
peak_counter = 0
thre1 = 0.1
thre2 = 0.05
for part in range(18):
map_ori = self.avg_heatmap[:, :, part]
one_heatmap = gaussian_filter(map_ori, sigma=3)
map_left = np.zeros(one_heatmap.shape)
map_left[1:, :] = one_heatmap[:-1, :]
map_right = np.zeros(one_heatmap.shape)
map_right[:-1, :] = one_heatmap[1:, :]
map_up = np.zeros(one_heatmap.shape)
map_up[:, 1:] = one_heatmap[:, :-1]
map_down = np.zeros(one_heatmap.shape)
map_down[:, :-1] = one_heatmap[:, 1:]
peaks_binary = np.logical_and.reduce(
(one_heatmap >= map_left, one_heatmap >= map_right,
one_heatmap >= map_up, one_heatmap >= map_down,
one_heatmap > thre1))
peaks = list(
zip(np.nonzero(peaks_binary)[1],
np.nonzero(peaks_binary)[0])) # note reverse
peaks_with_score = [x + (map_ori[x[1], x[0]], ) for x in peaks]
peak_id = range(peak_counter, peak_counter + len(peaks))
peaks_with_score_and_id = [
peaks_with_score[i] + (peak_id[i], )
for i in range(len(peak_id))
]
all_peaks.append(peaks_with_score_and_id)
peak_counter += len(peaks)
# find connection in the specified sequence, center 29 is in the position 15
limbSeq = [[2, 3], [2, 6], [3, 4], [4, 5], [6, 7], [7, 8], [2, 9],
[9, 10], [10, 11], [2, 12], [12, 13], [13, 14], [2, 1],
[1, 15], [15, 17], [1, 16], [16, 18], [3, 17], [6, 18]]
# the middle joints heatmap correspondence
mapIdx = [[31, 32], [39, 40], [33, 34], [35, 36], [41, 42], [43, 44],
[19, 20], [21, 22], [23, 24], [25, 26], [27, 28], [29, 30],
[47, 48], [49, 50], [53, 54], [51, 52], [55, 56], [37, 38],
[45, 46]]
connection_all = []
special_k = []
mid_num = 10
for k in range(len(mapIdx)):
score_mid = self.avg_paf[:, :, [x - 19 for x in mapIdx[k]]]
candA = all_peaks[limbSeq[k][0] - 1]
candB = all_peaks[limbSeq[k][1] - 1]
nA = len(candA)
nB = len(candB)
if nA != 0 and nB != 0:
connection_candidate = []
for i in range(nA):
for j in range(nB):
vec = np.subtract(candB[j][:2], candA[i][:2])
norm = math.sqrt(vec[0] * vec[0] + vec[1] * vec[1])
vec = np.divide(vec, norm)
startend = list(
zip(
np.linspace(candA[i][0],
candB[j][0],
num=mid_num),
np.linspace(candA[i][1],
candB[j][1],
num=mid_num)))
vec_x = np.array([
score_mid[int(round(startend[I][1])),
int(round(startend[I][0])), 0]
for I in range(len(startend))
])
vec_y = np.array([
score_mid[int(round(startend[I][1])),
int(round(startend[I][0])), 1]
for I in range(len(startend))
])
score_midpts = np.multiply(
vec_x, vec[0]) + np.multiply(vec_y, vec[1])
score_with_dist_prior = sum(score_midpts) / len(
score_midpts) + min(
0.5 * image.shape[0] / norm - 1, 0)
criterion1 = len(np.nonzero(
score_midpts > thre2)[0]) > 0.8 * len(score_midpts)
criterion2 = score_with_dist_prior > 0
if criterion1 and criterion2:
connection_candidate.append([
i, j, score_with_dist_prior,
score_with_dist_prior + candA[i][2] +
candB[j][2]
])
connection_candidate = sorted(connection_candidate,
key=lambda x: x[2],
reverse=True)
connection = np.zeros((0, 5))
for c in range(len(connection_candidate)):
i, j, s = connection_candidate[c][0:3]
if i not in connection[:, 3] and j not in connection[:, 4]:
connection = np.vstack(
[connection, [candA[i][3], candB[j][3], s, i, j]])
if len(connection) >= min(nA, nB):
break
connection_all.append(connection)
else:
special_k.append(k)
connection_all.append([])
# last number in each row is the total parts number of that person
# the second last number in each row is the score of the overall configuration
subset = -1 * np.ones((0, 20))
candidate = np.array(
[item for sublist in all_peaks for item in sublist])
for k in range(len(mapIdx)):
if k not in special_k:
partAs = connection_all[k][:, 0]
partBs = connection_all[k][:, 1]
indexA, indexB = np.array(limbSeq[k]) - 1
for i in range(len(connection_all[k])): # = 1:size(temp,1)
found = 0
subset_idx = [-1, -1]
for j in range(len(subset)): # 1:size(subset,1):
if subset[j][indexA] == partAs[i] or subset[j][
indexB] == partBs[i]:
subset_idx[found] = j
found += 1
if found == 1:
j = subset_idx[0]
if subset[j][indexB] != partBs[i]:
subset[j][indexB] = partBs[i]
subset[j][-1] += 1
subset[j][-2] += candidate[
partBs[i].astype(int),
2] + connection_all[k][i][2]
elif found == 2: # if found 2 and disjoint, merge them
j1, j2 = subset_idx
membership = ((subset[j1] >= 0).astype(int) +
(subset[j2] >= 0).astype(int))[:-2]
if len(np.nonzero(membership == 2)[0]) == 0: # merge
subset[j1][:-2] += (subset[j2][:-2] + 1)
subset[j1][-2:] += subset[j2][-2:]
subset[j1][-2] += connection_all[k][i][2]
subset = np.delete(subset, j2, 0)
else: # as like found == 1
subset[j1][indexB] = partBs[i]
subset[j1][-1] += 1
subset[j1][-2] += candidate[
partBs[i].astype(int),
2] + connection_all[k][i][2]
# if find no partA in the subset, create a new subset
elif not found and k < 17:
row = -1 * np.ones(20)
row[indexA] = partAs[i]
row[indexB] = partBs[i]
row[-1] = 2
row[-2] = sum(
candidate[connection_all[k][i, :2].astype(int),
2]) + connection_all[k][i][2]
subset = np.vstack([subset, row])
# delete some rows of subset which has few parts occur
deleteIdx = []
for i in range(len(subset)):
if subset[i][-1] < 4 or subset[i][-2] / subset[i][-1] < 0.4:
deleteIdx.append(i)
subset = np.delete(subset, deleteIdx, axis=0)
# subset: n*20 array, 0-17 is the index in candidate, 18 is the total score, 19 is the total parts
# candidate: x, y, score, id
return candidate, subset
@staticmethod
def __normalize(img, img_mean=(128, 128, 128), img_scale=1 / 256):
img = np.array(img, dtype=np.float32)
img = (img - img_mean) * img_scale
return img
@staticmethod
def __pad_width(img, stride, min_dims, pad_value=(0, 0, 0)):
h, w, _ = img.shape
h = min(min_dims[0], h)
min_dims[0] = math.ceil(min_dims[0] / float(stride)) * stride
min_dims[1] = max(min_dims[1], w)
min_dims[1] = math.ceil(min_dims[1] / float(stride)) * stride
pad = []
pad.append(int(math.floor((min_dims[0] - h) / 2.0)))
pad.append(int(math.floor((min_dims[1] - w) / 2.0)))
pad.append(int(min_dims[0] - h - pad[0]))
pad.append(int(min_dims[1] - w - pad[1]))
padded_img = cv2.copyMakeBorder(img,
pad[0],
pad[2],
pad[1],
pad[3],
cv2.BORDER_CONSTANT,
value=pad_value)
return padded_img, pad
def __call__(self, image):
return self.__get_auto_grading_outputs(image)
# Optimizes a model
# Relies on the platform => Must be compiled for every device
# The first command lin parameter must be the path to the old model
import sys
import torch
# import torch2trt
from models.with_mobilenet import PoseEstimationWithMobileNet
from torch2trt.torch2trt import *
MODEL_NAME = sys.argv[1]
NEW_MODEL_NAME = MODEL_NAME.replace(".pth", "_opt.pth")
WIDTH = 224
HEIGHT = 224
net = PoseEstimationWithMobileNet()
net.load_state_dict(torch.load(sys.argv[1]), strict=False)
net.eval().cuda()
data = torch.ones((1, 3, HEIGHT, WIDTH)).cuda()
model_trt = torch2trt(net, [data], fp16_mode=True, max_workspace_size=1 << 25)
torch.save(model_trt.state_dict(), NEW_MODEL_NAME)
def convert_to_skelets(in_, out_, cpu=False, height_size=256):
# height_size - network input layer height size
# cpu - True if we would like to run in CPU
print('start convert to skelets')
# mask that shows - this is bed
mask = cv2.imread(os.path.join('mask', 'mask.jpg'), 0)
mask = cv2.normalize(mask,
None,
alpha=0,
beta=1,
norm_type=cv2.NORM_MINMAX,
dtype=cv2.CV_32F)
net = PoseEstimationWithMobileNet()
load_state(net, checkpoint)
net = net.eval()
if not cpu:
net = net.cuda()
stride = 8
upsample_ratio = 4
max_number = 963
num_img = 0
stream = cv2.VideoCapture("rtsp://*****:*****@62.140.233.76:554")
# for num in range(0, max_number + 1):
while (True):
# frame = 'frame' + str(num) + '.jpg'
# img = cv2.imread(os.path.join(in_, frame), cv2.IMREAD_COLOR)
r, img = stream.read()
# cv2.destroyAllWindows()
# find the place of the bed - and add border to it, so we can cut the unnecessary part
# apply object detection and find bed
# output is an image with black pixels of not bed, and white pixels of bed
heatmaps, pafs, scale, pad = infer_fast(net, img, height_size, stride,
upsample_ratio, cpu)
total_keypoints_num = 0
all_keypoints_by_type = []
for kpt_idx in range(18):
total_keypoints_num += extract_keypoints(heatmaps[:, :, kpt_idx],
all_keypoints_by_type,
total_keypoints_num)
pose_entries, all_keypoints = group_keypoints(all_keypoints_by_type,
pafs)
for kpt_id in range(all_keypoints.shape[0]):
all_keypoints[kpt_id, 0] = (all_keypoints[kpt_id, 0] * stride /
upsample_ratio - pad[1]) / scale
all_keypoints[kpt_id, 1] = (all_keypoints[kpt_id, 1] * stride /
upsample_ratio - pad[0]) / scale
# how many persons in image
num_persons = len(pose_entries)
# num_img more than time_period - we delete first second and add the last second
bones_detected = np.zeros(len(bones_to_detect))
bones_xa = np.zeros(len(bones_to_detect))
bones_ya = np.zeros(len(bones_to_detect))
bones_xb = np.zeros(len(bones_to_detect))
bones_yb = np.zeros(len(bones_to_detect))
bones_in_bed = np.zeros(len(bones_to_detect))
for n in range(num_persons):
count_person_not_in_bed = 1
for id_x in range(len(bones_to_detect)):
bones_detected[id_x] = 0
bones_xa[id_x] = 0
bones_ya[id_x] = 0
bones_xb[id_x] = 0
bones_yb[id_x] = 0
bones_in_bed[id_x] = 0
if len(pose_entries[n]) == 0:
continue
for id_, part_id in enumerate(bones_to_detect):
kpt_a_id = BODY_PARTS_KPT_IDS[part_id][0]
global_kpt_a_id = pose_entries[n][kpt_a_id]
kpt_b_id = BODY_PARTS_KPT_IDS[part_id][1]
global_kpt_b_id = pose_entries[n][kpt_b_id]
# if both points are detected
if global_kpt_a_id != -1 and global_kpt_b_id != -1:
bones_xa[id_], bones_ya[id_] = all_keypoints[
int(global_kpt_a_id), 0:2]
bones_xb[id_], bones_yb[id_] = all_keypoints[
int(global_kpt_b_id), 0:2]
if mask[int(bones_ya[id_])][int(
bones_xa[id_])] == 1 and mask[int(
bones_yb[id_])][int(bones_xb[id_])] == 1:
bones_in_bed[id_] = 1
bones_detected[id_] = 1
sum_bones = 0
for id_, val in enumerate(bones_in_bed):
sum_bones += val
if sum_bones == len(bones_in_bed):
# anomaly
# we take mean vector of 2 vectors of bones 6 and 9
bone_xa = (bones_xa[0] + bones_xa[2]) / 2
bone_ya = (bones_ya[0] + bones_ya[2]) / 2
bone_xb = (bones_xb[0] + bones_xb[2]) / 2
bone_yb = (bones_yb[0] + bones_yb[2]) / 2
x1 = bone_xb - bone_xa
y1 = bone_yb - bone_ya
x2 = 100
y2 = 0
global anomaly_checker
alfa = math.acos(
(x1 * x2 + y1 * y2) /
(math.sqrt(x1**2 + y1**2) * math.sqrt(x2**2 + y2**2)))
# if alfa is close to 90 degree - anomaly
if min(abs(alfa - rad_90), abs(alfa - rad_270)) <= threshold:
print('num_persons', num_persons)
if num_persons == 1:
anomaly_checker = np.delete(anomaly_checker, 0)
anomaly_checker = np.append(anomaly_checker, 1)
cv2.imwrite(os.path.join('out_out', frame), img)
if np.sum(anomaly_checker) >= SEC_WITHOUT_HELP:
print('ALARM!')
num_img += 1
if not os.path.exists(out_):
os.mkdir(out_)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
print('done convert to skelets')
